GrayM Project

This Project is in development. It's not completed.

The GrayM Project aims to simulate the main parts of the brain and their functionalities using Python, with the ultimate goal of creating an Artificial Intelligence brain. The project is organized into different modules, each representing a distinct part of the brain.

Related Projects

https://github.com/0xroyce/Yolo-v8-face-detection-and-training

https://github.com/0xroyce/yolov8-dct

Project Structure

brain_project/
│
├── cerebrum/
│   ├── __init__.py
│   ├── frontal_lobe.py
│   ├── parietal_lobe.py
│   ├── temporal_lobe.py
│   └── occipital_lobe.py
│
├── cerebellum/
│   ├── __init__.py
│   └── cerebellum.py
│
├── brainstem/
│   ├── __init__.py
│   ├── midbrain.py
│   ├── pons.py
│   └── medulla_oblongata.py
│
├── thalamus/
│   ├── __init__.py
│   └── thalamus.py
│
├── hypothalamus/
│   ├── __init__.py
│   └── hypothalamus.py
│
├── limbic_system/
│   ├── __init__.py
│   ├── hippocampus.py
│   └── amygdala.py
│
├── basal_ganglia/
│   ├── __init__.py
│   └── basal_ganglia.py
│
├── memory/
│   ├── __init__.py
│   └── memory_storage.py
│
├── image_storage/
│   ├── __init__.py
│   └── image_handler.py
│
├── main.py
└── README.md

Module Responsibilities

Cerebrum

The cerebrum is divided into four lobes, each responsible for different high-level brain functions.

Frontal Lobe (cerebrum/frontal_lobe.py)
- Reasoning
- Planning
- Problem-solving
- Voluntary movement control
Parietal Lobe (cerebrum/parietal_lobe.py)
- Processing sensory information (touch, temperature, pain)
- Spatial orientation
Temporal Lobe (cerebrum/temporal_lobe.py)
- Auditory processing
- Memory
- Speech comprehension
Occipital Lobe (cerebrum/occipital_lobe.py)
- Visual processing

Cerebellum (`cerebellum/cerebellum.py`)

Coordination of voluntary movements
Balance and posture
Speech coordination

Brainstem

The brainstem connects the brain to the spinal cord and manages basic life functions.

Midbrain (brainstem/midbrain.py)
- Vision
- Hearing
- Eye movement
- Body movement
Pons (brainstem/pons.py)
- Breathing control
- Communication between brain parts
- Sensations (hearing, taste, balance)
Medulla Oblongata (brainstem/medulla_oblongata.py)
- Autonomic functions (breathing, heart rate, blood pressure)

Thalamus (`thalamus/thalamus.py`)

Relay station for sensory and motor signals to the cerebral cortex
Regulation of consciousness, sleep, and alertness

Hypothalamus (`hypothalamus/hypothalamus.py`)

Temperature regulation
Hunger and thirst control
Sleep and emotional activity regulation
Controls the pituitary gland (link between nervous and endocrine systems)

Limbic System

The limbic system manages emotions and memory.

Hippocampus (limbic_system/hippocampus.py)
- Formation of new memories
- Learning processes
Amygdala (limbic_system/amygdala.py)
- Emotion regulation (fear and pleasure responses)

Basal Ganglia (`basal_ganglia/basal_ganglia.py`)

Regulation of voluntary motor movements
Procedural learning
Routine behaviors or habits

Memory (`memory/memory_storage.py`)

Stores and retrieves learned data
Uses SQLite for data persistence

Image Storage (`image_storage/image_handler.py`)

Stores images for future recognition
Uses OpenCV for image recognition
Uses Pillow for image handling

Getting Started

Clone the repository:
```
git clone <repository-url>
```
Run the main script:
```
python main.py
```
Explore the modules:
- Each module can be extended with additional functions to simulate more detailed behaviors of the respective brain parts.
Add your Google Cloud JSON file to cerebrum/ folder
- Generate your Google Cloud JSON file for Speech to Text and Text to Speech and add it to /cerebrum folder. Then update temporal_lobe.py with name of your file.

GrayM uses YOLOv8 for visual recognition. Follow the steps below to set up and train the recognition engine.

Capture Images

Run capture_images.py to capture 300 images of the subject. The script captures an image every 0.2 seconds by default. You can modify the interval in the capture_images.py script if needed.

python cerebrum/capture_images.py

Instructions: Press 'c' to capture an image, and 'q' to quit. Enter the object's name when prompted. The captured images and metadata will be saved in the captured_images directory. 2. Prepare Dataset

Run prepare_dataset.py to organize the captured images and generate the necessary YOLO format labels.

python cerebrum/prepare_dataset.py

Instructions: This script copies the images to the yolo_dataset/images directory and creates placeholder labels in the yolo_dataset/labels directory. It also generates the data.yaml file required for YOLO training.

3.Train YOLO Model

Run train_yolo.py to train the YOLO model using the prepared dataset.

python cerebrum/train_yolo.py

Instructions: Ensure you have the ultralytics library installed. This script will train the YOLOv8 model and save the trained model in the trained_model directory.

4.Use the Recognise Engine

After training the model, you can use it for real-time object recognition. Run occipital_lobe.py to start the webcam and recognize objects using the trained model.

python cerebrum/occipital_lobe.py

Instructions: The script will load the trained YOLO model and start the webcam. Recognized objects will be displayed with bounding boxes and class names on the video feed.

Contributions

Contributions are welcome! Feel free to submit issues or pull requests to improve the project.

License

This project is licensed under the MIT License. See the LICENSE file for details.

0xroyce/GrayM